The present invention relates to additive manufactured parts and more specifically to predicting potential failure modes in additive manufactured parts.
Additive manufacturing is a process in which three-dimensional parts are formed on a layer-by-layer basis. Additive manufacturing can be used in place of other manufacturing methods like casting, forging and machining. Additive manufacturing can also be used to form parts having fine geometric features that are difficult to produce by the other methods described above.
In some additive manufacturing methods, beds of powder are used to create the layers of a part. A first powder bed is deposited according to a three-dimensional model and a heat source melts at least a portion of the first powder bed to form a starting layer of the part. The heat source can be a laser or an electron beam. An additional powder bed is then deposited on the first powder bed and the heat source melts a portion of the additional powder bed to form an additional layer of the part that is joined with the starting layer of the part. This process is continued on a layer-by-layer basis until the final part geometry is achieved.
The beds of powder and the part generated from the beds of powder are typically housed within a sealed chamber with a controlled atmosphere during the additive manufacturing process to reduce oxidation in the beds of powder. Because the additive manufacturing process typically occurs within a sealed chamber, monitoring the part for possible defects as the part is manufactured is difficult as there is limited space and access within the sealed chamber for sensors. Furthermore, the part is substantially covered by the beds of powder throughout the manufacturing process, obstructing the part from visual inspection as the part is manufactured.